SDSignal Desk

Benjamin Riley argues AI rests on a flawed model of the mind, and schools pay the price

· Updated · Signal Desk newsroom

Analysis of reporting by The Verge. How we work

Image: The Verge

In The Verge, education writer Benjamin Riley argues the brain-as-computer metaphor misleads AI builders and calls AI tempting in the moment but harmful as a steady diet for learners, citing student-usage claims, homework evidence and new bans in Norway, Los Angeles and New York City.

Key takeaways

  • Riley draws on neuroscientist Paul Cisek’s feedback-control view of the brain.
  • He calls AI tempting short-term and harmful as a steady diet, especially for learners.
  • He cites OpenAI’s claim that learners are more than half of ChatGPT users.
  • Norway has banned it for students under 13; Los Angeles and New York City districts have restrictions.
  • Riley discloses informally advising Schools Beyond Screens, which pushed for bans.

The argument

Writing in The Verge, education writer Benjamin Riley argues the AI industry works from a poor model of the human mind. The brain-as-computer metaphor of input, computation and output shapes how labs build and sell AI. Riley draws on neuroscientist Paul Cisek’s view that brains are better understood as feedback-control systems shaped by evolution and by social learning and institutions such as schools; Cisek has a forthcoming book that Riley says will expand on the model. The essay also cites François Chollet’s description of AI as cognitive automation and a scholarly argument that LLMs are better seen as cultural and social technologies than as intelligent agents.

His conclusion is blunt: AI is tempting in the moment and harmful as a steady diet, especially for learners. He points to OpenAI’s claim that learners make up more than half of ChatGPT’s users, Anthropic’s Claude for Teachers initiative, evidence that students stop doing homework once they start using AI, and new bans, including Norway for students under 13 and districts such as Los Angeles and New York City. Riley discloses that he has informally advised Schools Beyond Screens, a group that pushed for school chatbot bans.

Where the thesis overreaches and where it lands

Calling AI junk food for the mind flattens a tool that, used well, can explain a concept many ways, give a stuck student a next step or help a teacher with 150 students. Many people use these systems to learn more. A blanket diagnosis that the technology is inherently harmful is no better supported than the hype.

Riley is right on the central point. Learning requires effort, and people take off-ramps when offered. The essay cites researchers describing cognitive delegation: as people hand off reasoning, the social environment supporting independent thinking weakens, making delegation easier still. That is a credible feedback loop, and a frictionless homework machine invites it. A tool that does the work is a substitute, not a tutor. Presenting students as ChatGPT’s largest user group as a success could equally be a warning sign if much of that use is skipping the hard part.

What useful classroom AI would look like

The answer is neither pretending AI is absent from children’s lives nor flooding classrooms with free premium chatbots. Useful school AI is designed around friction: it asks questions back, declines to hand over finished essays and gives teachers visibility. Labs seeking credibility in education could measure learning outcomes rather than engagement and publish results. If usage remains the target, bans are likely to keep spreading. Pressure on schools comes from both sides, with vendor offers of free access and union calls and district bans. Related work on how AI fits into learning includes OpenAI’s teen initiatives (teen planning tools). More in our future of work coverage, research coverage and AI coverage.

Who feels it

Students
Risk skipping effortful practice; benefits appear only with tools built to support thinking.
Teachers and school leaders
Pulled between free AI offers and union calls and bans.
AI labs in education
Published learning-outcome evidence, not engagement, will decide the debate.

What to watch

  1. More districts or states restricting chatbots.
  2. Independent learning-outcome studies from OpenAI or Anthropic.
  3. Tutor modes that withhold direct answers, and whether students use them.
  4. Paul Cisek’s forthcoming book.

FAQ

What is Riley’s main argument?

That AI is built on a flawed brain-as-computer model and harms learners as a steady diet.

Whose neuroscience does he cite?

Paul Cisek’s feedback-control view of the brain.

Which bans does he mention?

Norway for students under 13, plus districts such as Los Angeles and New York City.

Does Riley have a stake?

He discloses informally advising Schools Beyond Screens, which pushed for bans.

Read the original

Continue at the source.

The Verge

Companies: Google

More on Google

All Google coverage →